OUP user menu

Lewy Body Pathology Involves Cutaneous Nerves

Masako Ikemura MD, PhD, Yuko Saito MD, PhD, Renpei Sengoku MD, Yoshio Sakiyama MD, PhD, Hiroyuki Hatsuta MD, Kazutomi Kanemaru MD, PhD, Motoji Sawabe MD, PhD, Tomio Arai MD, PhD, Genta Ito MS, Takeshi Iwatsubo MD, PhD, Masashi Fukayama MD, PhD, Shigeo Murayama MD, PhD
DOI: http://dx.doi.org/10.1097/NEN.0b013e318186de48 945-953 First published online: 1 October 2008

Abstract

Involvement of the peripheral autonomic nervous system is a core feature of Lewy body (LB) diseases, including Parkinson disease (PD), PD with dementia, and dementia with LBs. To investigate the potential use of skin biopsy for the diagnosis of LB diseases, we assessed anti-phosphorylated α-synuclein immunoreactivity in peripheral nerves in samples of skin from the abdominal wall and flexor surface of the upper arm in 279 prospectively studied consecutively autopsied patients whose data were registered at the Brain Bank for Aging Research between 2002 and 2005. Positive immunoreactivity was demonstrated in the unmyelinated fibers of the dermis in 20 of 85 patients with LB pathology in the CNS and the adrenal glands, the latter representing a substitute for peripheral autonomic nervous system sympathetic ganglia; no reactivity was seen in 194 patients without CNS LB pathology. In 142 retrospectively studied patients autopsied from 1995 onward who had subclinical or clinical LB disease, the sensitivity of the positive skin immunoreactivity was 70% in PD and PD with dementia and 40% in dementia with LBs. Skin immunoreactivity was absent in cases of multiple-system atrophy, progressive nuclear palsy, and corticobasal degeneration. We demonstrate for the first time that the skin is involved and may be a highly specific and useful biopsy site for the pathological diagnosis of LB diseases.

Key Words
  • α-Synuclein
  • Adrenal gland
  • Dementia with Lewy bodies
  • Dermis
  • Immunohistochemistry
  • Lewy bodies
  • Parkinson disease

Introduction

Lewy body (LB) diseases (LBDs) (1) are defined by neuronal degeneration related to the presence of LBs; they include Parkinson disease, either with normal cognition (PD) or with dementia (PDD), dementia with LBs (DLB), and LB-related progressive autonomic failure (LBPAF) (2). Involvement of the peripheral autonomic nervous system (PANS) is a key feature of LBD and is a presenting clinical feature in some cases of PD, PDD, DLB, and LBPAF. Autonomic dysfunction can greatly influence the patient's prognosis and quality of life.

Pathological studies of the PANS demonstrate LB pathology involving the sympathetic and enteric nervous systems (3-8). Recently, 123I-metaiodobenzylguanidine cardiac scintigraphy has become widely accepted in Japan as a tool for the diagnosis of LBD (9, 10); the pathological basis of this test is the presence of LB pathology of the sympathetic nerve fibers that supply epicardial fatty tissue (11).

There is a general consensus among neuropathologists that the standard organ for pathological evaluation of the PANS in LBD is the sympathetic ganglion (12), but autopsy sampling of the sympathetic ganglia can be difficult. We recently reported that the adrenal gland-one of the organs routinely examined at general autopsy-can be used as a substitute for the PANS sympathetic ganglia because it has similar pathological findings in LBDs (2).

The anatomical structures identified to date as having LB pathology are not appropriate biopsy sites for the premortem diagnosis of LBDs. In 2002, we examined skin excised near a decubitus ulcer at autopsy from a patient with DLB and found positive immunoreactivity with anti-phosphorylated α-synuclein antibodies in an unmyelinated fiber in the dermis. This observation prompted us to examine skin sampled from consecutively autopsied patients whose data were registered in the Brain Bank for Aging Research (BBAR).

Materials and Methods

Tissue Sources

The study consisted of 2 parts. The first part was a prospective study to determine the specificity of LBD in the skin in relation to LBD in the CNS and PANS. The second part was a retrospective study to determine the sensitivity of detection of LBD in the skin in relation to LBD in the CNS and PANS. The data registered in BBAR were from consecutively autopsied patients from a general geriatric hospital; informed consent was obtained from the relatives at autopsy (2, 13, 14).

In the prospective study, we used routinely sampled abdominal skin and prospectively sampled brachial skin from 279 consecutive autopsy patients whose data were registered in BBAR between 2002 and 2005. The patients' ages ranged from 52 to 104 years (mean, 80.8 ± 8.6 [SD] years); the male-to-female ratio was 167:112. The postmortem interval ranged from 52 minutes to 88 hours (mean, 13 hours). This series included 8 patients with progressive supranuclear palsy and 3 patients with corticobasal degeneration.

A wedge-shaped brachial skin sample, 1 cm × 0.5 cm in area and including the dermis and subcutaneous fatty tissue, was directly fixed in 4% paraformaldehyde for 48 hours and then embedded in paraffin. The abdominal skin had been routinely sampled since 1995, fixed in 10% buffered formalin for at least a week, and then embedded in paraffin. We used fixation in 4% paraformaldehyde for 48 hours because this fixation increased sensitivity for Lewy neurite detection in the CNS. We also continued to fix abdominal skin in 10% buffered formalin because the fixation is generally accepted and more applicable to potential biopsy in the clinic.

In the retrospective study, from among 1,594 patients whose data had been consecutively registered in BBAR between 1995 and 2005, we used 142 cases with the postmortem diagnosis of CNS LB stage II or greater (see later). The results of abdominal skin samples from 33 patients, who had been registered from 2002 onward and were also used in the prospective study, were included in the retrospective study to increase the case number for statistical analysis. Archival paraffin blocks of abdominal skin from the additional 109 patients from the period between 1995 and 2001 were stained as in the prospective study. The age range of the 142 patients was from 48 to 100 years (mean, 83.7 ± 7.6 years); the male-to-female ratio was 72:70. Abdominal skin samples from 3 cases with multiple-system atrophy (MSA; 1 case of MSA-P and 2 cases of MSA-C) from this period were also examined for comparison because no patient of MSA was included in the prospective studies. This study was approved by the institutional review boards of the Tokyo Metropolitan Institute of Gerontology and the Tokyo Metropolitan Geriatric Hospital.

Clinical Information

Clinical data, including information on the presence or absence of parkinsonism and cognitive state, were obtained from medical charts, as previously reported (13, 14). Final locomotive activity was classified into 4 levels: bedridden, wheelchair-bound, cane-assisted, and independent walking. In addition, the presence of decubitus ulcers was noted from the prosecutors' records at autopsy.

Pathological Examination of the Skin

Six-micrometer-thick serial paraffin sections of the skin were stained with hematoxylin and eosin and by immunohistochemistry with an autoimmunostainer (20NX; Ventana, Tucson, AZ) for single or double immunolabeling, as previously reported (15). All antibodies used are listed in Table 1.

View this table:
TABLE 1.

For LB pathology, the antibodies used were anti-phosphorylated α-synuclein (psyn) (psyn no. 64 16 monoclonal and PSer129 17 polyclonal). One section each of 2 serial sections was used for the staining of monoclonal and polyclonal anti-psyn antibodies. In addition, selected sections were double-immunostained with psyn and anti-phosphorylated neurofilament monoclonal antibody (SMI31).

Double labeling immunofluorescence studies were performed by incubating sections with anti-psyn antibody (PSer129, polyclonal) and anti-tyrosine hydroxylase (TH) antibody; for labeling, anti-rabbit Alexa 546 Fluor (Molecular Probes, Eugene, OR) (red) and immunoglobulin G Alexa 488 (green) were used. Sections were viewed under a Zeiss confocal laser scanning microscope (model LSM5 PASCAL; Jena, Germany).

For immunoelectron microscopy, we used selected paraffin sections stained with PSer129 and visualized using diaminobenzidine. The sections were evaluated by light microscopy and washed in a 0.1% (pH 7.4) phosphate buffer, postfixed in 1% osmium tetroxide, dehydrated in a graded series of ethanol, and embedded in epoxy resin. Plastic capsules filled with fresh epoxy resin were placed on the appropriate area of the sections and after polymerization of the resin, the target tissues were stripped off from the glass slides to the top of the epoxy resin. Ultrathin sections were then obtained and were examined under an H-7500 electron microscope (Hitachi, Japan) without a counterstain.

The BBAR Protocols for Evaluating Pathology of the CNS

The brains and spinal cords from all the patients of both the retrospective and the prospective studies were examined as previously reported (14). Briefly, 6-μm-thick sections were stained with hematoxylin and eosin and by the Klüver-Barrera method; selected sections were further examined with modified methenamine and Gallyas-Braak silver staining for senile changes, Congo red for amyloid deposition, and elastica Masson trichrome stain for vascular changes. In addition, selected sections were stained using the anti-α-synuclein and anti-ubiquitin antibodies (Table 1). To evaluate other senile changes, antibodies against phosphorylated tau, amyloid β (Aβ), glial fibrillary acidic protein, CD68, and phosphorylated neurofilament were also used.

Lewy body-related pathology in the CNS was examined at several levels of the thoracic spinal cord and in the medulla oblongata at the level of the dorsal motor nucleus of the vagus, the upper pons at the level of the locus ceruleus, the midbrain, the cerebellum (including the dentate nucleus), the basal ganglia (including the basal nucleus of Meynert), the amygdala, and the posterior hippocampus. In addition, areas were evaluated for LB scores according to the first and revised consensus guidelines for DLB (18, 19). These areas included the anterior cingulate gyrus, the entorhinal cortex, the second frontal and temporal gyri, and the supramarginal gyrus. Samples from these areas were stained by immunohistochemistry and classified into 7 CNS LB stages according to previously reported criteria (2, 14, 16) as follows: LB stage 0, no LBs; LB stage 0.5, Lewy neurites alone or diffuse or fine granular cytoplasmic staining lacking any focal aggregates in sections stained with anti-psyn antibodies; LB stage I, scattered LBs without cell loss (incidental LBD); LB stage II, abundant LBs with macroscopic loss of pigmentation in substantia nigra and locus ceruleus and/or gliosis demonstrated by glial fibrillary acidic protein immunohistochemistry in areas containing LBs but without attributable parkinsonism or dementia (subclinical LBD); LB stage III, PD without dementia; LB stage IV, DLB or PDD, transitional (limbic) form (DLBT or PDDT); and LB stage V, DLB or PDD, neocortical form (DLBN or PDDN). Parkinson disease with dementia was differentiated from DLB based on the definition in the consensus guidelines that “[PDD] dementia appears more than 12 months after the onset of parkinsonism.” The CNS LB stages are shown in Table 2. We subcategorized CNS stages I and II into primary and secondary α-synucleinopathy on the basis of our previous work (14, 16). Primary α-synucleinopathy involved the intermediolateral column of the spinal cord or the preganglionic sympathetic neurons and was further subdivided into brainstem, transitional, and neocortical forms according to the LB score or distribution. Secondary α-synucleinopathy spared the preganglionic sympathetic neurons, preferentially involved the amygdala, and was termed amygdala variant.

View this table:
TABLE 2.

Evaluation of Pathology Related to Other Senile Changes

Neurofibrillary tangles were classified into 7 stages and senile plaques into 4 stages according to Braak criteria (20). Argyrophilic grains were classified into the 4 stages that we previously reported (13). National Institute on Aging-Reagan criteria modified by us were adopted for the diagnosis of Alzheimer disease (AD) (21). Diagnoses of “dementia with grains” and “neurofibrillary-tangle-predominant form of dementia” were based on Jellinger's definitions (22, 23).

Pathological Study of the PANS

The adrenal glands were evaluated in the 279 patients in the prospective study, as previously reported (2). The adrenal glands from the additional 109 patients in the retrospective study were also examined for adrenal LB pathology. Data on 47 of these 109 patients were included in our previous report (2).

Statistical Analysis

Statistical analysis was performed by the χ2 test for comparisons of categorical data. Statistical significance was established at p < 0.05.

Results

Prospective Study

LB Pathology in the Skin

Immunohistochemical staining with anti-psyn antibodies demonstrated positive neurites and dots in nerve fascicles of the dermis and subcutaneous tissue (Figs. 1A, B). In some cases, psyn-positive nerve fibers showed swellings (Figs. 1C, D). Psyn-positive small dots or thin linear structures were also found around blood vessels (Figs. 1E, F). These psyn-positive structures seemed to colocalize with the axons. This was confirmed by double label immunohistochemistry using antibodies to psyn and phosphorylated neurofilament (Figs. 2A, B). By confocal microscopy, psyn immunoreactivity was also colocalized with the epitope of anti-TH antibody (Fig. 3).

FIGURE 1.

Immunohistochemical staining of nerve fascicles and vascular walls in the skin with anti-phosphorylated α-synuclein (psyn) antibodies. (A) Dotlike immunoreactivity (arrowhead) in a cross section of a nerve fascicle. (B) Longitudinal section of a cutaneous nerve with threadlike psyn immunoreactivity with a focal swelling (arrowhead). (C) Oval areas of immunoreactivity (arrowheads) scattered in nerve fascicles. (D) An oval structure connected to a linear structure (arrowhead) and small dotlike and linear staining (arrow). (E) Polyclonal anti-psyn antibody reveals a thin linear structure (arrowheads), extending from a nerve fascicle to the wall of blood vessel. (F) Monoclonal anti-phosphorylated α-synuclein antibody detects several positive thin linear structures in a vessel wall (arrowheads). Inset shows a higher power. (A-E): PSer129, polyclonal; and (F): psyn no. 64, monoclonal. Scale bars = (A-F) 25 μm; (inset in F) 10 μm.

FIGURE 2.

Double immunocytochemical staining with anti-phosphorylated α-synuclein (brown) and anti-phosphorylated neurofilament (red) antibodies. (A) In a cross section, PSer129 immunoreactivity is localized within SMI31-immunoreactive axons (arrowhead). (B) In a longitudinal section of a nerve fascicle, PSer 129 immunoreactivity is continuous with SMI31-immunoreactive axons (arrowheads). Scale bars = (A, B) 10 μm.

FIGURE 3.

Immunohistochemical and confocal immunofluorescent localization of phosphorylated α-synuclein (PSer129) and tyrosine hydroxylase (TH). (A) An oval area of immunoreactivity to PSer129 in the dermal nerve. Scale bars = 25 μm. (B-D) Confocal images of dermal nerve showing colocalization. (B) Alexa 546 (red) for phosphorylated α-synuclein. (C) Alexa 488 (green) for TH. (D) View of merged (B and C). Scale bars = (B-D) 10 μm.

By electron microscopy of immunostained sections, psyn-immunoreactive structures seemed to be composed of granulofilamentous profiles among mitochondria and vesicles. The psyn-immunoreactive structures had thin filaments at their peripheries that were consistent with the size of intermediate filaments and were surrounded by basal lamina. These features are consistent with those of unmyelinated axons (Fig. 4).

FIGURE 4.

Immunoelectron microscopy of anti-phosphorylated α-synuclein antibody (PSer129)-immunoreactive structures in a dermal nerve fascicle. (A) Light microscopy of an immunoreactive oval profile visualized with diaminobenzidine. Scale bars = 25 μm. (B) Immunoelectron microscopy showing the structure shown in (A) surrounded by basal lamina (triple arrows) and with abundant organelles. This probably represents the cytoplasm of an unmyelinated Schwann cell (arrowheads). Scale bars = 1 μm. (C) High-magnification view of the area indicated by the asterisk in (B), demonstrating granulofilamentous profiles among mitochondria and vesicular structures. Scale bars = 250 nm. (D) High-magnification view of the area indicated by the arrows in (B) showing thin filaments. Scale bars = 500 nm.

Comparison of CNS LB Stages

Psyn-positive structures in the skin were very small and not always found in the 2 serial sections stained either with psyn no. 64 and PSer129. Therefore, when positive immunoreactivity was found in any of the 2 stained sections for each anatomical location, the case was counted as positive. Correlations between CNS LB stage in the prospective study and the presence or absence of α-synuclein immunoreactivity in the skin are summarized in Table 3. Psyn immunoreactivity in abdominal and brachial skin did not always match in a single patient (Table 3), but the overall frequencies were similar and were apparently unaffected by differences in the method of fixation or in anatomical location.

View this table:
TABLE 3.

The numbers of patients and rates of skin positivity (%) at each stage were as follows: stage 0, none of 194 patients (0%); stage 0.5, 1 (4.0%) of 25 patients; stage I, 1 (3.7%) of 27 patients; stage II, 3 (33.3%) of 9 patients; stage III, 2 (100%) of 2 patients; stage IV, 6 (54.5%) of 11 patients; and stage V, 7 (63.6%) of 11 patients. Because all CNS LB stage 0 patients, including 8 progressive supranuclear palsy and 3 corticobasal degeneration cases, were negative for psyn immunoreactivity in the skin, there were no false-positive results, and the specificity was 100%. Among the 20 patients with psyn-positive immunoreactivity in the skin, 11 had positive immunoreactivity in both skin locations, 6 in the abdominal skin alone, and 3 in the skin of the arm alone.

Comparison With Adrenal Gland Positivity

All the patients with positive psyn immunoreactivity in the skin also had α-synucleinopathy in the adrenal glands (Table 3).

Retrospective Study

Pathology of LB-Related α-Synucleinopathy in the Skin

The numbers of patients and the rates of skin positivity (percentage) at each CNS LB stage of II or more were as follows: stage II, 13 (23.2%) of 56 patients; stage III, 10 (71.4%) of 14 patients; stage IV, 17 (44.7%) of 38 patients; stage V, 18 (52.9%) of 34 patients (Table 4). Samples from all 3 of the patients with MSA were negative.

View this table:
TABLE 4.

We next examined the percentages of patients with positive psyn immunoreactivity in the skin from the abdomen in each subgroup of CNS LB stage II or subclinical LBD (Table 4). The sensitivity was approximately 20%, but was 0% for the amygdala variant cases. We then classified stage IV and V cases into PDD and DLB and compared the positivity ratios. The number of patients and positivity ratio (%) in each group were as follows: PDDT, 8 (61.5%) of 13 patients; DLBT, 9 (36%) of 25 patients; PDDN, 6 (85.7%) of 7 patients; and DLBN, 12 (44.4%) of 27 patients. Therefore, the sensitivity of LB pathology in the skin was 14 (70%) of 20 in PDD and 21 (40.4%) of 52 in DLB.

Association of Coexistent Pathology of AD or Argyrophilic Grain Disease With LB-Related Skin Pathology in PDD or DLB

Patients with PDD or DLB were classified into 2 groups: those having or not having LB-related pathology in the skin. These 2 groups were evaluated for the grade of coexistent AD or argyrophilic grain disease (AGD) pathology. Nine (25.7%) of 35 PDD/DLB cases with psyn immunoreactivity in the skin were complicated by AD pathology or Braak neurofibrillary tangle stage equal to or more than III and senile plaque stage equal to C (21). In contrast, 14 of 37 anti-psyn negative PDD/DLB cases (37.8%) were complicated by AD pathology (Table 5). Seven (20%) of 35 PDD/DLB cases with LB pathology of the skin were complicated by AGD stage II or greater. Twelve (32.4%) of 37 cases without LB pathology in the skin were complicated by AGD stage II or greater. Thus, AD pathology or high-grade AGD pathology complicated 14 (40%) of 35 psyn-positive cases and 25 (67.6%) of 37 psyn-negative cases. The rate of complication by AD and/or AGD pathology was significantly higher in PDD/DLB patients without dermal α-synucleinopathy than in PDD/DLB patients with accompanying LB-related pathology in the skin (p = 0.019; Table 5).

View this table:
TABLE 5.

Clinicopathologic Correlation of Subclinical and Clinical LBD With LB Pathology in the Skin

Activities of daily living (ADL) data on the 84 patients without dermal LB-related pathology in the retrospective study were as follows: bedridden, 39 patients (46.4%); wheelchair-bound, 6 patients (7.1%); cane-assisted, 13 patients (15.5%); and walking independently, 26 patients (31.0%). The ADL data on the 58 patients with dermal LB-related pathology were as follows: bedridden, 37 patients (63.8%); wheelchair-bound, 7 patients (12.1%); cane-assisted, 4 patients (6.9%); and walking independently, 10 patients (17.2%). The proportion of bedridden patients was significantly higher (p < 0.001) and the proportion of patients walking independently tended to be lower (p = 0.065) among psyn-positive patients compared with psyn-negative patients. The numbers of bedridden patients with decubitus ulcers were as follows: dermal psyn-negative patients, 14 (35.6%) of 39; dermal psyn-positive patients, 21 (56.8%) of 37. Thus, patients with positive dermal LB pathology had a significantly higher rate of decubitus ulcers than dermal LB-negative patients (p = 0.028).

Discussion

This study is the first to demonstrate that LB-related pathology involves the cutaneous nerves in LBD. Because LB-related pathology identified in the skin was only seen in cases that had LB-related pathology in the CNS and not in 194 cases without CNS LB-related pathology, there were no false-positives; the specificity is, therefore, 100%. The sensitivity of LB pathology in the skin was 70% in PD and 40% in DLB in single sampling from the abdominal surface. The PDD or DLB cases without LB-related pathology in the skin were more frequently complicated by AD or AGD pathology than were those with LB-related pathology in the skin.

Sympathetic nerve fibers innervate blood vessels, eccrine glands, and the erector pili muscles in the skin. Sympathetic nerve fibers that innervate eccrine glands are cholinergic nerves, whereas all others are adrenergic. The psyn-positive structures were colocalized with axons visualized with SMI31 in the skin and were associated with axons by electron microscopy and TH immunohistochemistry. In light of these findings, we speculate that the psyn immunoreactivity was localized in the adrenergic nerves that innervate blood vessels and erector pili muscles in the skin. Not all psyn immunoreactivities were colocalized with anti-TH-immunoreactive axons, however, and it is more probable that α-synuclein also accumulates in the cholinergic nerves that innervate eccrine glands.

By electron microscopy, classic perikaryonal and intraneuritic LBs in the CNS are composed of a dense central core and surrounded by radiating filaments and vesicles. The LBs in the axons consist of randomly arranged or aggregated filamentous material with an inner core composed of electron-dense granular material. Immunohistochemical studies with anti-psyn antibodies have demonstrated axonal swellings in the cerebral white matter (16); these axonal swellings contained mitochondria and dense and lamellated bodies ultrastructurally (data not shown). Our electron microscopic findings of psyn-immunoreactive structures in the skin seem to be similar to those of axonal swellings in the CNS. This suggests that the cutaneous nerves were directly affected by LB-type accumulation of psyn.

Tests such as the thermal sweat test, pilocarpine sweat test, and sympathetic skin response have been used clinically to evaluate sudomotor function. There have been some clinical reports of the use of these tests to gauge sudomotor function in PD, but the severity and distribution of sudomotor dysfunction differ between each study (24-26). Studies suggest that the central or preganglionic sympathetic fibers may be involved in the early-stage lesions responsible for sudomotor dysfunction; postganglionic sympathetic fibers may become involved with progression of the disease. There have been few pathological studies of the skin in LBD (27), and no reports have described LB-related pathology in the skin. Our studies provide definite morphological evidence of the involvement of postganglionic axon terminals in LBD.

There is controversy regarding the distribution of clinical sudomotor dysfunction in LBD. Most reports indicate that although sweating preferentially decreases in the distal extremities, the distribution of the sudomotor dysfunction is patchy and differs among patients. Our prospective studies of the brachial and abdominal skin showed that psyn immunoreactivity was not always the same in skin from 2 different anatomical locations. Because the distal extremities were frequently affected by ischemic changes in our elderly cohort, we chose a proximal extremity and the trunk for the evaluation of LBD pathology. If skin biopsy is to become a useful tool in the diagnosis of LBD, it will be necessary to choose an area of skin where physiological tests consistently show definite abnormalities. Low sensitivity (about 20%) of dermal LB pathology among subclinical LBD cases (our CNS LB stage II) may also indicate that random skin biopsy is not valid for the early diagnosis of LBDs.

We retrospectively investigated possible correlations between the patients' ADL and skin LB pathology and found that ADL criteria were worse in patients with positive LB-related pathology in the skin than in those without it. This result is consistent with the clinical consensus in LBD that autonomic dysfunction usually parallels motor signs. Moreover, dermal psyn-positive patients had a significantly higher rate of complications in the form of decubitus ulcers than did the negative patients. Although the main cause of decubitus ulceration relates to motor disturbance in LBD patients, the direct involvement of the skin by LBD may be a contributing factor that warrants additional attention.

Our results also reveal that the epitope of phosphorylated α-synuclein accumulates in the cutaneous nerves in subclinical LBD. This finding is consistent with cardiac 123I-metaiodobenzylguanidine scintigraphic finding of involvement of unmyelinated fibers in the epicardial fatty tissue in incidental LBD (28).

Dermal psyn immunoreactivity was observed most frequently in PD (without dementia) and reached 100% in the prospective study, although the number of cases was small. In CNS LB stages IV and V, LB-related pathology in the skin was more frequent in PDD than in DLB. In addition, PDD and DLB cases without cutaneous LB-related pathology were more frequently complicated by the pathology of AD or AGD. Braak et al (29) proposed a staging paradigm whereby LB-related pathology first occurs in the dorsal motor nucleus of the vagus nerve, extends rostrally in the brainstem, reaches the limbic system, and eventually affects the cerebral cortex. Cumulative evidence indicates that LB-related pathology preferentially localizes itself in the amygdala in AD (30) and other tauopathies (14); this pattern of distribution has been termed the amygdala variant (2). We previously found that the adrenal glands were always free of LB-related pathology in such cases of amygdala variant (2). Although in the present study the number of cases was small, the skin was always free of LB-related pathology in the cases of amygdala variant studied (Table 4). These data further support our hypothesis that the presence of other dementing pathology in the CNS, such as AD or AGD, may induce cerebrally predominant distribution of LB pathology with relative sparing of the PANS.

One of the 279 patients in the prospective study presented with many LB-related pathological features in the abdominal skin and LBs in the adrenal glands but only very few Lewy neurites and dots in the dorsal motor nucleus of the vagus in the CNS. This case might represent the earliest stage of LBPAF.

For the differential diagnosis among parkinsonian syndromes, we additionally examined 3 more cases with MSA (2 MSA-P and 1 MSA-C patients) since 2006 and confirmed negative results. Although more cases should be analyzed for further confirmation and assessment of specificity, anti-psyn immunoreactivity of the skin was only positive for LBD and negative for MSA, progressive supranuclear palsy, and corticobasal degeneration in this study.

In conclusion, we document LB-related pathology involvement in cutaneous nerves in LBD. Skin biopsy may, therefore, have diagnostic application in cases of PD or LBPAF with advanced autonomic failure.

Acknowledgments

The authors thank Mr Naoo Aikyo, Ms Mieko Harada, Mr Satoru Fukuda, and Ms Nobuko Naoi (Department of Neuropathology, Tokyo Metropolitan Institute for Gerontology) for the preparation of sections and Dr Kinuko Suzuki for helpful discussions. We also thank 2 anonymous neurologists for preparing the Clinical Dementia Rating scale used in this study.

Footnotes

  • Masako Ikemura is now with the Department of Pathology, Teikyo University School of Medicine, Tokyo, Japan.

  • Grants-in-aid: Aid for Scientific Research on Priority Areas-Advanced Brain Science Project (SM, YuS) and Aid for Basic Scientific Research B (SM) and C (YuS) from the Ministry of Education, Culture, Sports, Science, and Technology of Japan; Aid for Degenerative Disease (SM), Neurological and Psychiatric Research (SM, YuS) and Research for Longevity (SM, YuS) from the Ministry of Health, Labor, and Welfare of Japan; Aid for Long-Term Comprehensive Research on Age-Associated Dementia from the Tokyo Metropolitan Institute of Gerontology (SM); Grant from The Novartis Foundation for Gerontological Research (2006) (YuS), Grant from the Tokyo Metropolitan Geriatric Hospital (YuS).

References

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
View Abstract